Sumo Robot Competition - Central Michigan Universitypeople.cst.cmich.edu/yelam1k/asee/proceedings/2016/student_regular_papers/2016_asee...Project Objectives The objective of this project

Proceedings of the 2016 ASEE North Central Section Conference

Copyright © 2016, American Society for Engineering Education

1

Sumo Robot Competition

Bryan Wilson

Student Author

Ohio Northern University

Ada, Ohio 45810

Email: [email protected]

Tyler Germann Student Author


Ada, Ohio 45810


Dr. Khalid Al-Olimat

Chair, Electrical & Computer Engineering and Computer Science Department

Student Group Adviser


Ada, Ohio 45810


Abstract

Sumo robot competitions have been prevalent in Japan, with the first competition being held in

1989 with only 33 robots, and its popularity has only been growing exponentially since its

inception. For example, in 2001, the number of robots entered into the Japanese competition was

over 4,000. In the early 1990s, the sumo robot competition was introduced into the United

States. Currently, most of the IEEE Student Activities Conferences (SAC) added the sumo robot

track to the list of the conference competitions.

There is a distinct lack in current literature (such as technical papers or conference proceedings)

regarding how to build a sumo robot. Thus, the purpose of this paper is to outline and explain the

design process of a robot that will compete at the IEEE Student Activities Conference (SAC)

Sumo Robot Competition tracks. Two different designs are explained; the first is utilizing both a

commercial off the shelf (COTS) kit along with custom chosen components, and the other design

utilizing only custom chosen components.

In addition to the design alternatives, the paper will discuss the decision matrix that includes the

factors and the criteria for the selection of each component and the final decision on which

design to go with for the competition. Discussions of the realistic constraints, the construction

phase, the testing, the budget, and the project schedule are included in the paper as well.

The authors strongly believe that this paper will help other students who are interested in

building a sumo robot and will save them a great deal of time in looking for material regarding

building a sumo robot.

Problem Statement

The Electrical & Computer Engineering and Computer Science Department (ECCS) at Ohio

Northern University would like a team of students to participate in the Sumo robot competition

at the IEEE Student Activities Conference. In addition, the department would like the robot to be

mailto:[email protected]





2

displayed and demonstrated to prospective students and their families during admission visitation

days.

Project Objectives

The objective of this project is to build a robot that serves two purposes: to be entered into the

Mini-Sumo Robot Competition and to be used as a showpiece for the Engineering College. In

order to be entered into the competition, all of the competition rules must be adhered to, which

includes constraints such as weight and cost. During the competition, this robot must be capable

of autonomously locating the opposing robot and pushing it out of the ring. Finally, the robot

should be durable enough to be demonstrated to prospective students during college events.

Literature Survey

Sumo robot competitions have been prevalent in Japan, with the first competition being held in

1989 with only 33 robots, and its popularity has only been growing exponentially since its

inception. For example, in 2001, the number of robots entered into the Japanese competition was

over 4,000, according to Pete Miles in his book that aims to serve as a general guide for building

sumo robots1. In the early 1990s, the sumo robot competition was introduced into the United

States, and a person named Bill Harrison quickly became its largest advocate. From there, he and

a colleague created the mini class of sumo robots (weighing in at 500g total) in order to invite

more people to design and compete, as it would be cheaper than the 3kg class of regular sumo

robots that are utilized in Japan1. Thus, the competition in the US was born.

Much like the regular sumo competitions (in which two opposing people aim to push each other

out of a ring), the robot variant of this event pits two robots against each other, each aiming to

push the other robot out of the ring, called a dohyo1. Each competition is governed by a set of

rules, which aim to keep the ensuing competition fair and exciting. For the mini-sumo class, here

are some important rules, found in the second reference2, which is the document that served as

last year’s sumo competition rules at The Ohio State University:

1. The mass limit of the mini-sumo robots is 500 grams (1.1 lbs)

2. The dimensions of the robot cannot exceed 10 cm by 10 cm (3.93 in by 3.93 in)

3. There is no limit to the height of the robot

These three rules serve as the guiding principle for which platform is chosen in a design. If any

of these requirements are not met, the robot will be disqualified. Each match of the tournament

consists of up to three rounds – the robot that wins two out of the three rounds will advance to

the next stage in the tournament. Effective points, known as yuko, are awarded in the following

cases1 :

1. when a robot pushes another robot out of the ring

2. when the opposing robot drives out of the ring

3. when the opposing robot is disqualified, or



3

4. when two yusei points are received.

To receive a yusei point, the opposing robot must get stuck on the border line and not be able to

move off of the line. Thus, to win a match, a team must receive two yuko points.

Because the basic idea is to push another robot out of the dohyo, there are various governing

scientific principles that must be considered in order to have an effective sumo robot design. In

summary, here are the important principles, according to1:

1. Output Torque: The robot must possess enough torque in order to be able to push the

mass of the other robot out of the ring (in most cases, more than just the mass alone, since

the other robot’s wheels will also be opposing the robot’s push). The relationship

between torque and applied force is shown in (1). F is the force in Newtons (N), T is the

torque in Newton-meter (N∙m), and r is the distance in meter (m) between the point that

the force is applied and the point of rotation:

𝐹 =𝑇

𝑟 (1)

As can be seen, if the torque is greater for a given lever arm, the force applied will be

greater. From this, it can be concluded that smaller wheels would be better – this is due to

the fact that given a specified torque, if the wheel is smaller, there will be a greater output

force. To increase or decrease torque, gear ratios can be utilized.

2. Pushing force: Because there is a specific coefficient of friction associated with materials

on other materials (specifically, types of wheels in contact with the dohyo), it follows that

there is then a force needed in order to overcome this coefficient of friction in order to

push something. Equation (2) illustrates this idea, where Ff is the frictional force in

Newtons (N), µ is the coefficient of friction, and FN is the normal force of the robot in

Newtons (N) (which is equal to the force of its weight):

𝐹𝑓 = 𝜇𝐹𝑁 (2)

The force exerted by the wheels and motors must be greater than this frictional force,

assuming the other robot is stationary.

3. Momentum: If the robot has a higher momentum, it is more likely to displace the other

robot, since an object with higher momentum is harder to resist/stop. Momentum

equation is presented in (3), where P is the momentum in Newton-seconds (N∙s), m is the

mass in kilograms (kg), and v is the velocity in meters per second (m/s):



4

𝑃 = 𝑚𝑣 (3)

Because velocity and torque are inversely proportional, there must be a good balance

between these two items in order to have an effective robot.

Overall, these equations denote the basic theory behind the sumo robot. If these ideas are taken

into consideration effectively, then the robot possesses a better chance of performing admirably.

Current implementations have some differences and many similarities. However, the overall

system of the sumo robot is the same – the following is a list of the necessary components for a

sumo robot, as highlighted in first reference1, along with some generalizations:

1. Sensors: Both line sensors and opponent detection sensors are needed. Line sensors are

utilized so that the robot does not drive out of the ring unintentionally, and opponent

sensors are utilized in order to determine in which direction to travel to attack the

opponent. Generally, the line sensors are phototransistors, which can sense reflectivity,

while the opponent sensors are either infrared (IR) or ultrasonic sensors.

2. Motors: Generally, there are two to four motors per mini-sumo robot. There are various

types of motors, but the two main types of motors utilized in this class of competition are

small DC motors and continuous rotation servo-motors.

3. Power: In general, in the mini-sumo class, most of the kits utilize four AA batteries in

order to power the entire device (which includes sensors, motors, microcontroller, etc.).

Some have rechargeable batteries, which are usually of the Lithium-Polymer (Li-Po)

variety. Any type of battery can be utilized as long as it can effectively power the robot

for an entire match.

4. Control: Overall, the prominent method of controlling the robot is by utilizing a

microcontroller such as an Arduino or a BASIC Stamp module. One design exception,

however, implemented control via only discrete components. In all methods, the overall

goal is to determine a control solution that can integrate with the sensors and motors in

order to cause the robot to function properly.

As a brief comparison, three different kits were taken into consideration as an initial starting

point from which to then create two custom designs – the Parallax SumoRobot3, the Solarbotics

Sumovore4, and the FingerTech Cobra chassis implementation5 (an image of each can be seen in

Appendix A). The Parallax robot has continuous rotation servo-motors, is powered by 4 AA

batteries, is controlled by a BASIC Stamp 2 module, has 2 line sensors, and 2 IR sensors for

opponent detection. The Solarbotics Sumovore is controlled with discrete components, is

powered by 4 AA batteries, has 2 DC motors, and has similar line and opponent sensing as the

Parallax robot. Finally, the FingerTech Cobra chassis has 4 DC motors, 3 line sensors, and is

powered by a rechargeable Li-Po battery (the other components are excluded – the designer

chooses opponent detection as well as the method for controlling the robot). As seen, these three



5

designs have both similarities and differences, which all factor into which design performs the

best.

Some limitations of these current designs include the method utilized for powering the actual

robot. The implementations with 4 AA batteries are cheap; however, these batteries, as noted in

reference 11, may not last for multiple matches. This causes them to be thrown away, thus

potentially impacting the environment. Also, because this robot is eventually going to be utilized

for demonstrations, it makes sense to have a solution that can be recharged, since that would

incur a smaller cost over time. Another limitation is the motors – because these robots are small,

the motors utilized have to be equally small. The continuous rotation servos do not provide much

mechanical advantage; however, they are very light. In comparison, the DC motors are heavier,

but have a higher mechanical output. Another drawback is that some of the DC motors then need

motor controllers, which can be realized in the form of solid state H-bridges1 – this incurs higher

cost in the end, but allows for more flexibility in utilizing DC motors.

The design concept that will be presented in this proposal is a general combination of different

pieces of current implementations. Because more mechanical power is needed, DC motors are

going to be utilized, while an implementation with a microcontroller will be presented, as it

allows for greater expandability. Finally, for powering the robot, a rechargeable battery will be

utilized for economic reasons.

As a side note, there is a distinct lack in current literature (such as technical papers or conference

proceedings) regarding how to build a sumo robot. Most of the information above was attained

from1, while different product websites were consulted for the kits listed earlier. Although this is

the case, there is some literature on how to program the sumo robots effectively. Two papers

found, reference6 and reference7, related to control functionality. In6, which is a paper by Hamit

Erdem, a fuzzy logic implementation is utilized in order to result in faster tracking of an

opponent, which then results in quicker target acquisition. In this scenario, the robot with a fuzzy

logic controller can respond faster and consequently attack the opposing robot. As a result, the

robot can respond more efficiently to the dynamic environment of combat. In7, which is a paper

taken from the publication Robotica, a multi-phase genetic programming approach is explored

for the control of the robot. According to the authors, genetic programming “applies genetic

manipulations to the functions and operators of a control program or other representations of a

controller in an autonomous system”7. In essence, the original functions are “mutated” by means

of parameters and values as “training” occurs. After enough training has been done, the robot

should be able to then respond to any given situation. Given the topics of the two papers as well

as the lack of literature on how to build the robot, this suggests that the majority of the time

should be spent on refining the algorithm that controls the functionality of the robot. Thus, when

determining the optimal design, this must be a consideration.

Realistic Constraints Performance:



6

The robot must detect the opposing robot if it is within 50 cm with 90% accuracy and a

response time of 10 ms or less.

The robot must detect the arena border with 100% accuracy to avoid round loss.

The robot should be less than or equal to the maximum weight limit of 1.1 lbs.

The robot should not exceed the dimension restrictions of 3.93” x 3.93” x h”

Functionality:

Vision sensors collect data that is sent to the microprocessor. A decision is then made by

the code and sent to the motors to correctly manipulate movement.

Line sensors collect data that is sent to the microprocessor. The decision to move away

from the line is sent to the motors to avoid round loss.

Economic:

The cost of the finalized robot must not exceed $250 as per the mini SumoBot rules.

The total cost of the project should stay around $500 as per the guidelines for the senior

design project.

Energy:

Minimum output voltage from power supply is 6V.

Minimum mAh battery’s rating should be enough to last for about three rounds.

Maintainability:

Check all bolts, screws, and nuts before operation and between rounds.

Batteries should be fully charged before use.

Batteries should not be left plugged into robot for prolonged durations.

The ECCS department would like to be able to keep the robots for display and demos so

a user instructions will be included.

Operational:

The robot must perform and interact with an opposing robot in a circular playing field

with a diameter of 77 cm.

The robot must detect the border of the arena to make sure it remains in play.

The robot must be able to run for the entire length of the match, 3 minutes.

Reliability:

The robot must operate for the entire length of the round, 3 minutes.

The robot must operate for a minimum of 4 rounds.

The robot should remain functional past the competition date for demos at ONU.

Availability:

The robot must be capable of working on demand, failure could result in loss of round at

the competition.



7

Alternative Solutions

To determine the best possible solution, three different designs are chosen for comparison – one of which is a kit, while the other two designs are custom. By

building upon a kit’s design and tweaking different subsystems slightly (such as the motors or the batteries), the optimal design can be achieved. The

determination of these optimal components is a direct result of the literature survey, since the information provided discussed the types of sensors that would

be best for this competition. Because the aim of the manufacturers is to decrease the end cost of their product (while still making profit), they chose

components that did not necessarily have the highest performance. Thus, by taking parts of these kits and adding in better components, a better design is

achieved, which can be seen below in Table 7.

Table 1: Decision Matrix

Components Weight Design1: Parallax Kit Design2: Cobra Design3: Vex Robot

Component Rating Score Component Rating Score Component Rating Score

Edge Detection 15 2 QTI sensors 0.2 3 3 QTI sensors 0.4 6 3 QTI sensors 0.4 6

Motors 10 2 Cont. Rotation

Servos 0.1 1 4 DC Motors (50:1) 0.45 4.5 2 Vex DC Motors 0.45 4.5

Opponent Detection 15 2 IR Rx/Tx 0.25 3.75 1 Sharp IR, 2 IR Rx/Tx 0.375 5.625 1 Sharp IR, 2 IR Rx/Tx 0.375 5.625

Battery 10 4 AA batteries 0.1 1 Rhino LiPo Battery, 360 mAh 0.4 4 2 Lithium Batteries 0.5 5

Mechanical Power 10 Average 0.1 1 Better 0.5 5 Better 0.4 4

Wheel Dimensions** 2.62in D, 0.3in W 3cm D, 2.16cm W 2.62in D, 0.3in W

Microcontroller 10 BASIC Stamp 2 0.45 4.5 BASIC Stamp 2 0.45 4.5 Arduino 0.1 1

Motor Control** None Two Dual H-Bridge None

Price** $129.99 $197.28 $175

Weight 10 0.67241 lbs 0.1 1 0.9061 lbs 0.4 4 1.050 lbs 0.5 5

Dimensions** Less than 3.93in x

3.93in 3.898in x 3.839in 3.90in x 3.90in

Software Development 20 Low Complexity 0.5 10 Mid Complexity 0.4 8 High Complexity 0.1 2

Center of Gravity** Mid Low Low

TOTAL 25.25 TOTAL 41.625 TOTAL 33.125

*** denotes that these are areas that were NOT considered in the weighting and point determination of the design, as some are restrictions placed upon each

design in order to be eligible for the competition (such as dimensions and price) while others are just other details used for comparison that do not affect the

robot’s eligibility. These aspects do not necessarily affect the overall robot functionality, but are good for informational purpose.



8

Design Solution

At this time, the current design that shall be entered into the competition is the Cobra design,

which is presented as Design 2 in Table 1 (refer to Appendix A for a figure). The Components

for this design are included in Table 2.

Table 2: Cobra Design

Component Quantity

BASIC Stamp 2 Sumo PCB 1

IR LED with Receiver 2

Sharp IR Sensor 1

Cobra Chassis 1

Rhino 360mAh, 11.1V 1

Breadboard Wires 30

The BASIC Stamp 2 Sumo PCB is taken from a kit that is offered by Parallax, and the board

includes components such as a voltage regulator, header pins (for adding extra accessories as

well as quicker prototyping), breadboard space, terminal for the power supply, dedicated

locations for the IR LEDs, and the BASIC Stamp 2 microcontroller (and the necessary

components for its operation, such as the crystal for the clock as well as a memory module). This

board is utilized in order to quickly prototype the overall design, as it offers many simple

connection points for sensor interfacing.

Also, the Cobra chassis, sold by FingerTech Robotics, includes all of the necessary components

to run four DC motors along with a microcontroller. Parts include: two dual H-bridge modules

for motor control, four 50:1 DC motors, three QTI line sensors, and four polyurethane wheels.

This particular configuration was chosen in order to have the best possible parts needed for

success – as an example, the four DC motors ensures that there is enough torque to be able to

push another robot out of the ring, even if the opponent is pushing against the Cobra head on. In

addition to this, the H-bridge motor control interfaces easily with the microcontroller, as these

components need to only have 5V applied for forward (a standard HI output level for a

microcontroller pin, especially the BASIC Stamp 2 module) and 0V applied for backward

movement (along with a PWM input signal, which the BASIC Stamp 2 module is capable of

doing). Finally, the line sensors allow the robot to detect the border, so again, the chassis is

convenient in that it allows for these components to already be factored into the design. The

entire design is powered by a 360mAh, 11.1V Li-Po battery, which allows for the robot to last

several rounds.

Finally, the different IR sensors are utilized for opponent detection. The IR LEDs return a 1 or 0

depending on whether or not the opponent is seen, while the Sharp IR sensor returns a

continuous voltage range dependent upon how far away the detected object is (which is

highlighted in the Literature Survey section). Ultrasonic sensors are not utilized due to possible

interference – if the opponent also uses ultrasonic, the sensors on the Cobra design would pick

up those signals, and thus be confused.

Design Integration



9

This project involves the implementation of two different designs, therefore this section will

discuss the process of designing and working on these prototypes. The first of the two designs,

the Cobra, is nearly complete at this stage. This design uses the cobra chassis kit and a Basic

Stamp 2 processor. At this moment the vision sensors used are included with the Basic Stamp

but will be changed out to a more reliable alternative. To begin, building the Cobra chassis was

quite straight forward, piecing together the components until the base was complete. Then the

basic stamp 2 was attached above the motors of the chassis using posts to leave room for the

battery compartment of the design. The most difficult part was wiring up the chassis connections

to the processor as there was a lack of clear documentation on the web. This was only a minor

setback and it was squared away and testing began. Finally, moving forward, the vision sensors

will be replaced and the processor will need to be attached in a different manner to accommodate

a slightly larger sensor.

The second design, the Vex bot, is not as far along in the design process as issues with the size

constraint are showing difficulty to work around. This design involves a small aluminum chassis

with two motors catty-corner to each other and four wheels to stabilize the base. This design is

still being tweaked but this will be the general layout of the robot. To see an initial physical

design of the Vex design, refer to Appendix A.

Design Testing and Verification

To test the Cobra design, it is being run against a standard Parallax kit that is used to give a

baseline for both designs. As only one design is currently available for testing this is what is

being focused on. The two robots are placed into the sumo arena and a round is played between

the two. At this stage, it is important to look at the big details such as tracking the opposing

robot and detecting the border. The Cobra robot still needs some work as it does not always

detect the edge, this is likely because of the test code that was taken from the FingerTech website

as it uses, by nature, very sequential programming (meaning that while the robot could be

turning to avoid running off the border, it will not care where the opponent is; it will simply

finish its motion no matter what). Moving forward changes will be made to the code to fix these

issues. By modifying different values and subroutines, the robot will be more likely to detect the

border and the component, allowing for better operation.

Perhaps the most important test is determining if the robot is within the weight restrictions, the

Cobra weighs in about 40 grams lighter than the weight limit which gives some wiggle room

when making any modifications. Equally as important is the size restrictions the robot must

meet. This was measured and the design fits the criteria perfects maximizing its size.

The last test that was conducted was the Cobra’s ability to push another sumo bot out of the ring.

The power difference between the Parallax kit and the Cobra design was quite noticeable and the

Cobra had no problem out powering the Parallax robot even in direct head-to-head collisions.

The overall test plan can be seen in the next section, which highlights the ongoing tests.

There are various tests that are to be done on the robot prior to the competition. By doing so,

there is a greater chance that the robot will perform reliably at the competition, thus allowing for



10

the highest possible chance of winning. The following list includes the necessary tests in order to

check for proper functionality:

1. To simulate a real competition scenario, two mini SumoBots will be designed and built to

test the strength of each design. The designs will be continuously modified based on their

performance. Before the competition date, these two robots will compete to see which

shall go to compete in the competition.

2. To check if the robot can push an opposing robot out of the ring, a 1.5 pound weight shall

be placed in the center of the ring. The robot must be able to push this weight out of the

ring.

3. In order to prove that the robot can effectively push out a 1.5 pound weight at any given

location in the ring, that weight shall be placed at different locations, and the robot will

be oriented in different positions, thus allowing the tracking system to find the mass and

consequently push it out of the ring.

4. To ensure that the motors are moving properly, connect only the motors and power

supply to the microcontroller and run a test program that will force the motors to run at

various speeds.

5. To test how long the robot will function, the robot will be run continuously inside of the

ring until the battery is fully drained. This will serve as a benchmark for determining how

long the battery lasts under that rated condition. The voltage of the power supply will be

monitored throughout this process, allowing for a better view into that particular

operating condition.

6. To test how long the robot will function in a competition setting, the robot will be

subjected to a stress test, in which the robot will be pushing up against an inanimate

object. By doing this, this forces the motors to stall, consequently increasing the amount

of current flowing from the power supply. In this scenario, the amount of time that the

robot can last while pushing a load heavier than it is capable of will show the operating

time of the robot if the maximum load is applied.

7. To test the line sensors, the microcontroller will be programmed such that it shall charge

up the capacitor in parallel with the phototransistor for a brief duration, turn off, and

consequently measure the time it takes for the capacitor to fully discharge. Since the

phototransistor acts as a variable resistor, the discharge time will be affected by the

amount of light that is sensed. For example, if the color white is detected by the

phototransistor, the capacitor will discharge slower – the microcontroller is able to then

measure and display that discharge time.

8. To ensure that the robot is within the mass limitations, weigh the robot with all

components included on the robot.

9. To ensure that the robot is within the size constraints, measure the robot’s width and

length.

10. To determine if the robot is ready for competition, test the fully built design by placing

other robots into the ring and have a mock tournament.

11. To test all of the software subroutines, set up each scenario that the program could

possibly go through, and test accordingly.

12. To ensure no crosstalk between sensors, ensure that the algorithm turns off all sensors

except the sensor that is being utilized.



11

Lessons Learned

Throughout the testing process, it was found that the design possibly warranted the use of more

sensors. Although software complexity is proportional to the number of sensors utilized, there is

an inherent complexity level that needs to be reached for the robot to actually have a chance of

being competitive during the competition.

The largest piece of information learned involved the engineering tradeoffs that were required to

be made. In general, the largest factor in all of the decisions involved the budget – because a

robot can only cost $250 total when competing, this constraint does not allow for many

expensive components. One might like to buy a very nice sensor; however, that leaves less

money for the motors, microcontroller, battery, and casing. Alternatively, one can purchase a

very nice microcontroller/microprocessor (in order to better implement a more involved

program, such as machine learning); however, that leaves less money for all of the other aspects

of the design.

Specifically towards the Sumo design, by testing out the Parallax kits, it was found that the IR

LEDs were not the best option, as their performance was not consistent. Due to this, the extra

sensors found through the literature survey are to be utilized, resulting in a more robust design.

This shows that different types of sensors are needed – both very simple and somewhat complex

(all staying within the cost constraints). The old saying of “not putting all of one’s eggs into the

same basket” certainly holds true – multiple sensors are needed for both redundancy as well as

for a competitive edge.

Design Budget Assuming that the Cobra design is utilized per the decision matrix, the following table shows the

cost breakdown for the components:

Table 3: Cost of Cobra SumoRobot Design

Component Price

Cobra Chassis w/Sensors $ 108.24

Rhino 3S Battery, 750mAh $ 9.79

Cobra PCB Mount Chassis $ 4.47

Parallax Sumo PCB $ 49.99

Sharp IR Sensors x 1 $ 19.99

LED Rx/Tx $ 4.80

TOTAL: $ 197.28

As can be seen, the total cost for the Cobra design costs $188.28, which is less than the $250

budget. The total seen in the above table is the cost for one SumoRobot, however, two robots

will be built for testing purposes as well as demonstrations at the university. For this reason the

total cost of components will be around $400.



12

Design Schedule

3-Sep 18-Sep 3-Oct 18-Oct 2-Nov 17-Nov 2-Dec 17-Dec 1-Jan 16-Jan 31-Jan 15-Feb 1-Mar 16-Mar 31-Mar 15-Apr

Written Proposal

Oral Proposal

Peer-Peer Evaluations

Preliminary builds

Written Progress Report

Order Parts

Oral Progress Report

Ethics Assignment

Rules Released

Start prototype

Programming

Testing

Redesign

Testing

Competition

SumoBot Gantt Chart



13

Conclusion

Overall, as seen throughout this paper, the proposed design successfully meets the criteria

needed for participating in the sumo robot competition. As such, the detailed process of research,

choosing components, building, and testing the robot are included within this paper, which

should serve as a guide for future students that would like to participate in this exciting event. By

doing so, one does not need to go searching for material around the Internet in order to find

resources on how to build a robot such as this – they need only to view this paper to gain the

insight needed to build a sumo robot (and, hopefully make an even better design!).

References

[1] P. Miles, Robot Sumo: The Official Guide, Berkeley: McGraw-Hill Osborne, 2002.

[2] A. Pang, "Sumo Robot Competition," 4 January 2015. [Online]. Available:

https://ieee.osu.edu/sac/competitions/sumo-robot-competition. [Accessed 27 September 2015].

[3] Parallax, "SumoBot Robot," Parallax, Inc., 13 January 2014. [Online]. Available:

https://www.parallax.com/product/27400. [Accessed 27 September 2015].

[4] Solarbotics, "Solarbotics Sumovore Mini-Sumo," Solarbotics, 12 March 2013. [Online]. Available:

https://solarbotics.com/product/k_sv/. [Accessed 27 September 2015].

[5] F. Robotics, "FingerTech 'Cobra' Mini-Sumo Chassis," FingerTech Robotics, 11 April 2014. [Online].

Available: http://www.fingertechrobotics.com/proddetail.php?prod=ft-kit-cobra-chassis. [Accessed 27

September 2015].

[6] H. Erdem, "A Practical Fuzzy Logic Controller For Sumo Robot Competition," in Pattern Recognition and

Machine Intelligence, Kolkata, Springer, 2007, pp. 217-225.

[7] J. Liu and S. Zhang, "Multi-Phase Sumo Maneuver Learning," Robotica, vol. 22, no. 1, pp. 61-75, 2004.



14

Appendix A: Diagrams of Different Designs

FingerTech Robotics Cobra Design5

Parallax SumoBot Kit3

Solarbotics Sumovore Kit4

Initial Vex Robot Design

Documents

Sumo Robot Competition - Central Michigan Universitypeople.cst.cmich.edu/yelam1k/asee/proceedings/2016/student_regular_papers/2016_asee...Project Objectives The objective of this project